Hello,
I am running a time-series regression where I have to split index returns in to positive and negative variables whilst keeping them on the same dates. If I split them up prior (through using excel) the regression doesn't want to run so I'm assuming there is a command to split them up throughout the time series. If anyone know how to perform this that would be great as I'm fairly new to eviews
change in VIX=bo + b1VIXt-1+ SPXret+ +SPXret-......
Splitting positive and negative values within same timeserie
Moderators: EViews Gareth, EViews Steve, EViews Moderator, EViews Jason
-
EViews Gareth
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13603
- Joined: Tue Sep 16, 2008 5:38 pm
Re: Splitting positive and negative values within same times
It isn't clear what you mean by this.
I'm assuming you have a series (a column of data) called SPXret. You have another column called VIX.
The question is, what do you want the values of SPXret+ to be when SPXret is negative? And what do you want the values of SPXret- to be when SPXret is positive?
I'm assuming you have a series (a column of data) called SPXret. You have another column called VIX.
The question is, what do you want the values of SPXret+ to be when SPXret is negative? And what do you want the values of SPXret- to be when SPXret is positive?
Re: Splitting positive and negative values within same times
the full regression i am trying to run is:
change in vix = bo + b1VIXt-1 + SPXret+ + SPXret- + DevMA5+ + DevMA5- + e
I have data from 2006 to 2010. I am looking at the effect contemporaneous index returns has on the VIX (looking at leverage effect). I have a full series of data for SPX returns but i need to separate them somehow in eviews to all the positive returns and all the negative returns. If i split them up beforehand i will only have roughly 2 half sets of data (positive and negative variables). this won't run on eviews as it says I obviously dont have enough data per variable. There must be a way of sorting positive and negative returns in to separate variables when running the regression. I don't want to have any other value for data that isn't in the correct column (eg, 0's when data isn't positive in SPXret+ variable). This has obviously been done as it has been shown in previous journals but the explanation as to how it was computed is very vague.
I hope this may clear things up and any help would be great
change in vix = bo + b1VIXt-1 + SPXret+ + SPXret- + DevMA5+ + DevMA5- + e
I have data from 2006 to 2010. I am looking at the effect contemporaneous index returns has on the VIX (looking at leverage effect). I have a full series of data for SPX returns but i need to separate them somehow in eviews to all the positive returns and all the negative returns. If i split them up beforehand i will only have roughly 2 half sets of data (positive and negative variables). this won't run on eviews as it says I obviously dont have enough data per variable. There must be a way of sorting positive and negative returns in to separate variables when running the regression. I don't want to have any other value for data that isn't in the correct column (eg, 0's when data isn't positive in SPXret+ variable). This has obviously been done as it has been shown in previous journals but the explanation as to how it was computed is very vague.
I hope this may clear things up and any help would be great
Re: Splitting positive and negative values within same times
I am not sure, but the following might help:
Code: Select all
vix c vix(-1) spxret*(spxret>0) spxret*(spxret<0) devma5*(devma5>0) devma5*(devma5<0)Who is online
Users browsing this forum: No registered users and 2 guests
